We describe the synthesis and characterization of a series of water-soluble free-base, zinc, and copper porphyrin–oligonucleotide (ODN) conjugates. A non-charged tetraarylporphyrin was directly attached to the 50-position of thymine via a short amide linker. Such a linker should allow for rigid connection to the adjacent nucleobases, thus increasing the sensitivity for monitoring conformational changes of the ODNs by electronic circular dichroism due to capping effects or ligand binding. Two complementary synthetic approaches have been used to incorporate porphyrin chromophores into the DNA. In the first approach a porphyrin carboxylic acid is conjugated to 50-amino-ODN. In the second approach the phosphoramidite of the porphyrin-amido-thymidine is coupled to 50-hydroxy-ODN using standard automated phosphoramidite chemistry. In both cases a spontaneous metallation of the free-base porphyrin in porphyrin– DNA conjugates was observed during the porphyrin–DNA conjugate cleavage from the solid support and its consequent deprotection using concentrated aqueous ammonia. Zinc and copper porphyrin– DNA conjugates were isolated by HPLC and characterized together with the anticipated free-base porphyrin– DNA conjugate. Authentic zinc and copper porphyrin–DNA conjugates were intentionally prepared from intentionally metallated porphyrins to compare their spectroscopic and HPLC characteristics with isolated metallospecies and to confirm the spontaneous metallation.

1. Synthesis and characterization of water-soluble free-base, zinc and copper
porphyrin–oligonucleotide conjugates
Angela Mammana a
, Tomohiro Asakawa a
, Klaus Bitsch-Jensen a
, Amanda Wolfe a
, Saireudee Chaturantabut a
,
Yuko Otani a
, Xiaoxu Li b
, Zengmin Li b
, Koji Nakanishi a
, Milan Balaz a,*, George A. Ellestad a,*, Nina Berova a,*
a
Department of Chemistry, Columbia University, 3000 Broadway, New York, NY 10027, USA
b
Columbia Genome Center and Department of Chemical Engineering, Columbia University, 1150 St. Nicholas Avenue, New York, NY 10032, USA
a r t i c l e i n f o
Article history:
Received 18 April 2008
Revised 7 May 2008
Accepted 9 May 2008
Available online 15 May 2008
Keywords:
Porphyrin
DNA
Oligonucleotide
Metallation
Phosphoramidite
MALDI-TOF
a b s t r a c t
We describe the synthesis and characterization of a series of water-soluble free-base, zinc, and copper
porphyrin–oligonucleotide (ODN) conjugates. A non-charged tetraarylporphyrin was directly attached
to the 50
-position of thymine via a short amide linker. Such a linker should allow for rigid connection
to the adjacent nucleobases, thus increasing the sensitivity for monitoring conformational changes of
the ODNs by electronic circular dichroism due to capping effects or ligand binding. Two complementary
synthetic approaches have been used to incorporate porphyrin chromophores into the DNA. In the first
approach a porphyrin carboxylic acid is conjugated to 50
-amino-ODN. In the second approach the phos-
phoramidite of the porphyrin-amido-thymidine is coupled to 50
-hydroxy-ODN using standard automated
phosphoramidite chemistry. In both cases a spontaneous metallation of the free-base porphyrin in por-
phyrin–DNA conjugates was observed during the porphyrin–DNA conjugate cleavage from the solid sup-
port and its consequent deprotection using concentrated aqueous ammonia. Zinc and copper porphyrin–
DNA conjugates were isolated by HPLC and characterized together with the anticipated free-base porphy-
rin–DNA conjugate. Authentic zinc and copper porphyrin–DNA conjugates were intentionally prepared
from intentionally metallated porphyrins to compare their spectroscopic and HPLC characteristics with
isolated metallospecies and to confirm the spontaneous metallation.
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
The ‘bottom-up’ approach to produce functional, self-organizing
nanodevices from molecular building blocks giving a precise
arrangement of atoms and molecules in three dimensions is most
likely to meet the needs of future technologies. Chromophore ar-
rays often exhibit different spectroscopic properties from mono-
mers that they are assembled from. Porphyrin assemblies have
been extensively investigated for their promising photonic proper-
ties. Many nature-inspired porphyrin assemblies have been devel-
oped aiming for efficient light collection and energy transfer.1–6
DNA has become very attractive as a molecular scaffold for pre-
cise positioning of desired molecules on the nanoscale due to its
molecular recognition and mechanical properties.7–9
Stacking
interactions between nucleobases are, in addition to hydrogen-
bonding, the key factors that stabilize the double-stranded DNA
helix and determine DNA’s secondary and tertiary structure.10,11
Exposed DNA termini base pairs represent the weakest link in
base-pairing fidelity and potential problems in the self-assembly
process. A well-defined arrangement of chromophores is an impor-
tant tool for novel applications in material sciences and nanotech-
nology. In order to increase the base pair fidelity at the DNA
termini, several synthetic strategies have been introduced. The
capping of the oligonucleotides termini with small molecules is a
useful method to improve the properties of their natural counter-
parts such as nucleobase pairing fidelity,12,13
duplex stability,14–17
or resistance against exonuclease degradation.18
Capping is usually
favored by covalently attaching small molecules to the 30
or 50
DNA
termini.
Porphyrins are one of the most studied DNA binding agents,
and there are many reports on porphyrin–DNA interactions.19–23
However, most studies focused on interactions between free
(non-covalently attached) cationic porphyrins and polymeric
DNA. Although the first porphyrin–oligonucleotide conjugates
were synthesized 15 years ago,24,25
there are only few recent
studies focusing on the structural and spectroscopic properties
of these conjugates,26–28
on interactions between porphyrins,29–31
and on interactions between porphyrin and DNA32,33
in porphy-
rin–DNA conjugates. We were the first to study the through space
exciton coupling between two porphyrins covalently attached to
0968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.05.041
* Corresponding authors. Present address: Oxford University, Department of
Chemistry, Mansfield Road, Oxford OX1 3TA, UK (M.B.). Tel.: +1 212 854 3934; fax:
+1 212 932 1289.
E-mail addresses: mb2402@columbia.edu (M. Balaz), gae2104@columbia.edu
(G.A. Ellestad), ndb1@columbia.edu (N. Berova).
Bioorganic & Medicinal Chemistry 16 (2008) 6544–6551
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc

2. a short DNA sequence (DNA as the helical scaffold), and the effect
of covalently attached porphyrins on DNA structure and stability.
We had reported previously that porphyrins linked to one of the
DNA termini can communicate through space upon formation of
double-stranded DNA resulting in an exciton-coupled circular di-
chroic signal.29
This signal is very diagnostic and can serve as
internal probes for DNA conformational changes.29,34
Recently
we have shown that a porphyrin, covalently attached to the 50
end of a short non-self-complementary oligonucleotide, can stabi-
lize the non-Watson–Crick base pairing between guanine and
adenine and promote the formation of the right-handed A-like
double-stranded DNA via termini capping.17
Additionally, porphy-
rin chromophores also served as a sensitive internal circular di-
chroic sensor of the newly formed secondary structure.17
Better
understanding of porphyrin–DNA interactions will help us to
rationally design new materials and predict their photophysical
and photochemical properties.
The position, length, and type of the linker between porphyrin
and DNA play an important role in the porphyrin/DNA interfacial
interaction and strongly dictate/influence the interactions between
the porphyrin and neighboring nucleobases. We anticipate that a
short and non-charged linker will promote a close non-covalent
stacking interaction between porphyrin and exposed DNA termini.
In this paper we present the synthesis of water-soluble 50
-por-
phyrin–DNA conjugates and 50
-metalloporphyrin–DNA conjugates
with a short 50
-amide linker between porphyrin chromophore and
the nucleobase. We have rationalized the origin of the unexpected
spontaneous metallation and discuss the different spectroscopic
and HPLC properties of the metalloporphyrin–DNA conjugates.
2. Results and discussion
2.1. Synthesis of porphyrin–DNA conjugates
Porphyrins were attached to the 50
position of self-comple-
mentary oligonucleotide sequences via a short, rigid amide link-
age. Annealing of the self-complementary porphyrin modified
ODN will provide a DNA duplex having one porphyrin cap at each
end. The amide bond was chosen because such a linkage would
provide a shorter and less flexible connection to the adjacent
nucleobases, thus increasing the sensitivity for monitoring con-
formational changes of the DNAs via long-range exciton-coupled
CD. Such a linker also provides good control over the position
of the porphyrin on the DNA and prevents unwanted porphyrin
intercalation. The short DNA sequence 50
-TGCGCGCA was selected
for our study. We used two different strategies for the synthesis
of the 50
-amido linked tetraarylporphyrin oligonucleotide. In the
first approach porphyrin carboxylate 1 is conjugated to solid sup-
ported 50
-amino-ODNs 4a–c (4a: hexamer, 4b: octamer, 4c:
decamer) using benzotriazol-1-yl-oxytripyrrolidino-phosphonium
hexafluorophosphate (PyBOP) and diisopropylethylamine (DIPEA,
Scheme 1). In the second approach the 50
-porphyrin-amido-thy-
midine 30
-phosphoramidite 9 is synthesized and then coupled to
50
-hydroxy-ODNs 10a–c under standard phosphylation conditions
(Scheme 3).
2.1.1. Approach 1
As outlined in Scheme 1, porphyrin carboxylic acid 1 was con-
jugated manually to the exposed 50
-amino-group of thymidine in
ODNs 4a–c. ODNs 4a–c were prepared from unmodified ODNs 50
-
hydroxy-(GC)nGCA-support 3a–c (n = 1–3) using standard method-
ologies on a DNA synthesizer. A 1000 ÅA
0
controlled pore glass (CPG)
solid-support cartridge was critical as the CPG cartridge with smal-
ler 500 ÅA
0
pores was much less successful for the attachment of the
bulky porphyrin carboxylic acid 1. The porphyrin–DNA conjugates
5a–c were then cleaved from the CPG beads and deprotected with
concentrated aqueous ammonia at room temperature overnight
(Scheme 2). The aqueous ammonia was removed under vacuum
and the crude reaction mixture analyzed by mass spectrometry
and reverse phase HPLC (RP-HPLC) to provide M-6a–c.
2.1.2. Approach 2
The synthesized 50
-amino-30
-hydroxy thymidine35,36
7 was re-
acted with the free-base trispyridylphenylporphyrin carboxylate37
1 (Scheme 3). The resulting 50
-porphyrin-amido-thymidine 8 was
converted to phosphoramidite 9, purified by flash column chroma-
tography, and characterized by 1
H and 31
P NMR. 31
P NMR showed
two signals resulting from two diastereomers around the pyrami-
dal phosphorous atom formed (Fig. 1). Porphyrin–phosphoramidite
9 was then reacted with solid supported oligonucleotides 10a–c
(10a: hexamer, 10b: octamer, 10c: decamer) under standard phos-
phoramidite chemistry. Concentrated aqueous ammonia was used
Scheme 1. Approach 1: synthesis of porphyrin–oligonucleotide conjugates 5a, 5b, and 5c Reagents: (i) standard automated cyanoethyl-phosphoramidite solid-phase
chemistry, DNA synthesizer; (ii) 50
amino-thymine-phosphoramidite; (iii) PyBOP, DIPEA, DMF, on-column synthesis.
A. Mammana et al. / Bioorg. Med. Chem. 16 (2008) 6544–6551 6545

3. for cleavage of the porphyrin–oligonucleotide conjugates 5a–c
from the solid support and DNA deprotection (Scheme 2). The
aqueous ammonia was removed under vacuum and the crude reac-
tion mixture analyzed by mass spectrometry and reverse phase
HPLC.
2.2. Analysis of porphyrin–DNA conjugates 6a, 6b, and 6c
2.2.1. MALDI-TOF and HPLC
MALDI-TOF analysis of the crude porphyrin-octamer 6b indi-
cated the presence of the expected free-base porphyrin–DNA 2H-
6b with a MW of 3051 ± 3. In addition, a species with a MW of
3114 ± 3 (3051 + 63) was observed, which seemed to be a metallo-
porphyrin-8-mer conjugate containing either copper(II) (+63.5) or
zinc(II) (+65.4) in the porphyrin coordination center. The HPLC pro-
file of 6b also showed two major peaks (Fig. 2). Porphyrin–DNA
conjugate 6b was dissolved in water and injected onto the HPLC
with a gradient of buffer (hexafluoroisopropanol and triethyl-
amine) and methanol at a flow rate of 1 mL/min (see Section
4.2.6. for further details).37
The first (more polar) compound eluted
at 47.1 min gave the anticipated MW of 3051 ± 3 for free-base
porphyrin–ODN 2H-6b. The second peak eluted at 48.6 min (less
Scheme 2. Cleavage from the solid support and the deprotection of the porphyrin–DNA conjugates 5a, 5b, and 5c.
Scheme 3. Approach 2: synthesis of porphyrin–oligonucleotide conjugates 5a, 5b, and 5c Reagents: (i) EDC, DCM; (ii) 2-cyanoethyl-N,N-diisopropylchlorophosphoramidite,
DIPEA, DCM; (iii) DCM, activation reagent (acetonitrile/tetrazole) then oxidation.
Figure 1. 31
P NMR spectrum of 50
-porphyrinthymidine 30
-phosphoramidite 9.
6546 A. Mammana et al. / Bioorg. Med. Chem. 16 (2008) 6544–6551

4. polar) gave a MW of 3114 ± 3. However due to the small mass dif-
ferences between zinc and copper (difference of two mass units)
MALDI-TOF spectrometry was unable to tell with certainty if we
had synthesized zinc porphyrin–ODN Zn-6b or copper porphy-
rin–ODN Cu-6b (together called M-6b). Similar results have been
obtained in the synthesis of porphyrin-hexamer 6a and porphy-
rin-decamer 6c. Two major products were always observed, the
first one corresponding to the desired free-base porphyrin–DNA
conjugates, and the second one to the metalloporphyrin-conju-
gates M-6, sometimes Zn-6 or Cu-6, and sometimes both. MAL-
DI-TOF data, HPLC retention times, and HPLC profiles for
porphyrin–DNA conjugates 6a, 6b, and 6c are summarized in Table
1, in Figure 2, and in Figure S1–S6 (see Supplementary material).
2.2.2. UV–vis absorption spectra
Because of the ambiguity of MALDI-TOF data we turned our
attention to absorption spectra of the porphyrin–ODNs 6a–c. It is
well known that the UV–vis profile of each porphyrin and metallo-
porphyrin is very diagnostic.38
UV–vis spectrum of first eluted free-
base porphyrin–DNA conjugate 2H-6 showed a Soret absorption
band at 422 nm and three Q bands at 516, 554, 585 nm. This spec-
tral profile is typical of the free-base porphyrins. Examination of
the UV spectrum of less polar M-6b (elution time 48.6 min) en-
abled us to assign this species to the zinc(II) porphyrin–DNA con-
jugate Zn-6b. The UV–vis spectrum porphyrin region showed a
maximum at 425 nm and two smaller Q bands at 558 and
594 nm characteristic of a zinc porphyrins.38
Surprisingly, using
different batches of ammonia, a copper porphyrin conjugate, Cu-
6b, was isolated with a Soret band at 418 nm and only one Q band
at 541 nm. Figure 3 shows UV–vis spectra of all three porphyrin–
ODNs, free-base conjugate 2H-6b, zinc conjugate Zn-6b, and cop-
per conjugate Cu-6b.
The HPLC analysis of porphyrin–DNA conjugates prepared via
approach 2 showed one major product with retention time of the
metallated porphyrin–DNA conjugate M-6b. The MALDI-TOF spec-
trum showed the molecular weight of 3114 (2H-6b + 63). As with
the results from Approach 1, we have explained the origin of these
unexpected compounds M-6b by the incorporation of copper(II)
(atomic weight = 63.5 g/mol) or zinc(II) (atomic weight = 65.4 g/
mol) into the porphyrin coordination center. The UV–vis spectrum
was also used to identify the presence of zinc and copper porphy-
rin–ODN conjugate M-6b.
2.3. Synthesis of metalloporphyrins M-1 and
metalloporphyrin–DNA conjugates M-6b
To confirm the origin of the metalloporphyrin–DNA conju-
gate(s), isolated along with the free-base porphyrin–DNA conju-
gate, we synthesized the zinc tetrapyridylporphyrin carboxylate,
Zn-1, and the corresponding Cu-1. Free-base porphyrin 1 was met-
allated using zinc acetate or copper acetate, respectively. Porphyrin
carboxylate 1 was dissolved in chloroform and 10 equivalents of
corresponding acetate was added. The reaction mixture had to be
heated to reflux as metallation was not observed at room temper-
ature. The synthesized metalloporphyrins Zn-1 and Cu-1 were
found to be of sufficient purity to be directly conjugated with 50
-
amino-ODN 4b attached to the solid support using identical condi-
tions as for free-base porphyrin 1. After cleavage and deprotection
with concentrated ammonia, porphyrin modified ODNs Zn-6b and
Cu-6b were purified by RP-HPLC (see Figures S8 and S9, Supple-
mentary material) and analyzed by UV–vis spectroscopy and MAL-
DI-TOF-MS. The UV–vis spectra were identical with those of
previously isolated zinc and copper porphyrin conjugates 6b
resulting from metal contamination. Although the free-base conju-
gate 2H-6b elutes typically earlier than the metal conjugates (see
Fig. 2), we observed negligible or no difference in retention time
between copper and zinc conjugates when authentic samples of
the last two were co-injected. Table 2 summarizes the UV–vis
characteristics of synthesized porphyrin-octamers M-6b.
2.4. Spontaneous metallation of porphyrin–DNA conjugates 6a,
6b, and 6c
We believe that the source of the metallation is the concen-
trated aqueous ammonia used for the cleavage/deprotection of
porphyrin–DNA conjugates. It is known that concentrated ammo-
nia contains numerous trace metals and that the long exposure of
the matrix during the ODN cleavage and deblocking of the assem-
bled conjugate gives enough time for metallation to take place.
That the concentrated ammonia is the source of the metals is fur-
Figure 2. HPLC profiles of crude porphyrin–DNA conjugates 6a (porphyrin-hexa-
mer, top), 6b (porphyrin-octamer, middle) and 6c (porphyrin-decamer, bottom)
monitored at two different wavelengths, 260 nm and 420 nm.
Table 1
HPLC and MALDI-TOF-MS data of products formed in the synthesis of porphyrin–DNA
conjugates 6a, 6b, and 6c
Porphyrin–DNA Calculated
molecular weight
HPLC retention
timea
(min)
MALDI MALDI
differenceb
6a (hexamer) 2434.6 51.2 2435.2 +0.6
52.8 2497.1 +62.5
6b (octamer) 3052.7 47.1 3054.51 +1.81
48.6 3112.0 +62.1
6c (decamer) 3670.8 43.5 3673.2 +2.4
44.9 3737.8 +67.0
a
See Table 3 for HPLC elution conditions.
b
Difference between calculated and experimental molecular masses.
A. Mammana et al. / Bioorg. Med. Chem. 16 (2008) 6544–6551 6547

5. ther corroborated by the fact that different ratios of the metallat-
ed species were obtained depending on the source of the aqueous
ammonia (results not shown). The small scale of the synthesis
(1 lM) and overall low yields of the porphyrin–DNA coupling
(<10%) undoubtedly contribute to the high level metallation as
well. The spontaneous metallation of porphyrin in ODN conju-
gates 6a, 6b, and 6c points to the need for caution in the use of
concentrated ammonia in the deprotection step involving
porphyrins.
It should be mentioned, however, that unconjugated free-base
porphyrin carboxylate 1 only underwent metallation with an ex-
cess of metal acetates ($10 equiv) under reflux (see below). The
mild experimental conditions under which we observe metal com-
plexation of the porphyrin conjugates are inconsistent with the
conditions necessary for complexation of free porphyrin 1. Based
on the MALDI-TOF-MS of crude porphyrin–ODN conjugate of dif-
ferent preparations, the amount of metalloconjugates present ran-
ged from 10% to 50% of the free-base porphyrin–ODN conjugates.
Additional porphyrin metallation takes place during the HPLC puri-
fication (50 °C) to give a mixture of unmetallated and metallated
porphyrin conjugates in a ratio of ca. 40/60, respectively, based
on HPLC product peak areas (see Fig. 2). Thus it appears that the
DNA plays an important role in porphyrin metal complexation,
which is further promoted during the HPLC step. The ability of
DNA to promote the metallation of porphyrins had been previously
observed by Li and Sen.39
3. Conclusions
In conclusion, we have synthesized a series of water-soluble
porphyrin–DNA and metalloporphyrin–DNA conjugates. Porphyrin
carboxylate 1, and Zn(II)- and Cu(II)-porphyrin derivatives have
been attached to the 50
position of thymine via a non-charged
amide bond applying two complementary synthetic strategies.
Spontaneous metallation of porphyrin in free-base porphyrin–
DNA conjugates 6 has been observed under mild conditions. We
believe that trace metals in the commercial, concentrated aqueous
ammonia are the origin of the unexpected metallation of the por-
phyrins. MALDI-TOF and UV–vis analysis have been employed to
identify the zinc and copper in metallated species. We anticipate
that the short non-charged amide linker together with different
metals in the coordination center of the porphyrin will exhibit
new spectroscopic and binding properties in porphyrin–DNA con-
jugates and help to further understand the physicochemical prop-
erties of porphyrin-capped DNA oligomers. Indeed, on-going UV,
CD, and fluorescence spectroscopic studies on these porphyrin–
DNA conjugates provide evidence for through space exciton cou-
pling, intermolecular porphyrin–porphyrin stacking, and porphy-
rin end-capping effects of the DNA oligomers as influenced by
environmental conditions. The results of these studies will be pub-
lished elsewhere.
4. Experimental
Reagents and materials were obtained from commercial suppli-
ers and used as received. Pre-dried dichloromethane (DCM) was
freshly distilled from CaH2 under argon before each use. If neces-
sary, the reactions were performed in vacuum dried glassware un-
der argon. Purification was performed by flash column
chromatography using Machery-Nagel silica gel (0.04–0.063 mm).
TLC was performed on Merck TLC plates silica gel 60 F254 utilizing
UV light (254 nm). ODNs attached to 1000 Å controlled pore glass
solid-support cartridges were prepared on an ABI-380 nucleic acid
synthesis system. DNA purification cartridges (Puri-Pak 0.2 lM),
activation reagent (0.45 M tetrazole/acetonitrile), and oxidation
solution (0.02 M iodine/pyridine/H2O/THF) were purchased from
ChemGenes Corporation.
The porphyrin–ODNs were purified on a Waters Delta HPLC
with Waters In-line Degasser AF system using a Waters reverse
phase X-Terra semipreparative column MS C18 2.5 lm
10 Â 50 mm at 50 °C, 1 mL/min flow. Evaporation of fractions from
HPLC was done using a CentriVap Concentrator with a CentriVap
Cold Trap.
Mass spectra were measured on a JMSHX110/110 Tandem mass
spectrometer (JEOL, Tokyo, Japan) under Fast Atom Bombardment
(FAB) 10 kV Accel-volt conditions with meta-Nitro Benzyl Alcohol
(m-NBA) as the matrix. MALDI-TOF-MS were performed on a Voy-
ager Applied Biosystems spectrometer using 3-HPA and diammo-
nium citrate in 50% MeCN in H2O. 400 MHz 1
H NMR, 75 MHz 13
C
NMR, and 162 MHz 31
P NMR spectra were obtained on a Bruker
Advance spectrometer and are reported in ppm (d) with coupling
constants in Hertz (Hz). 1
H spectra were referenced to different
internal standards, TMS, DMSO, and MeOH. 31
P spectra were refer-
enced to an external 85% aqueous H3PO4 standard. UV–vis spectra
were recorded on a JASCO V-530 spectrophotometer.
Figure 3. UV–vis absorption spectrum of porphyrin-8-mer conjugates 2H-6b (red), Zn-6b (blue), and Cu-6b (green). Soret band region (middle) and Q band region (right) of
UV–vis spectra were expanded. Conditions: 50 mM potassium phosphate buffer, pH 7.0.
Table 2
UV–vis absorption characteristics of porphyrin DNAs 2H-6b, Zn-6b, and Cu-6b
measured in 50 mM K-phosphate buffer, pH 7.0
Porphyrin-octamers Soret banda
Q band
2H-6b 422 nm (230,600) 516, 554, 585 nm
Zn-6b 425 nm (351,038) 557, 594 nm
Cu-6b 418 nm (251,546) 541 nm
a
Absorption wavelength (extinction coefficient).
6548 A. Mammana et al. / Bioorg. Med. Chem. 16 (2008) 6544–6551

6. 4.1. Synthesis of porphyrin–thymidine building blocks
4.1.1. 50
-Porphyrin-CONH-30
-hydroxythymidine (8)
Porphyrin carboxylate 1 (45 mg, 0.068 mmol) was dissolved un-
der argon in dry DCM (2 mL). To the purple solution, 50
-amino-30
-
hydroxythymidine (18.0 mg, 0.075 mmol), DMAP (16.6 mg,
0.136 mmol), HOBt (9.19 mg, 0.068 mmol), and EDC (26.1 mg,
0.136 mmol) were added. The solution was stirred at room temper-
ature for 3 h. The crude reaction mixture was loaded directly onto a
silica gel column and eluted using a mixture of DCM/MeOH 95:5
with 1% of triethylamine (TEA) to give 48.1 mg of 8. Porphyrin–thy-
midine 8 was then dissolved in 25 mL DCM and washed with water
to remove the triethylammonium salt impurity. The organic phase
was dried with MgSO4, filtered, and evaporated to dryness. 43.2 mg
(72%) of 8 as a purple powder was obtained. TLC (MeOH/DCM, 1:9)
Rf = 0.34. 1
H NMR (400 MHz, MeOH-d4): 1.91 (3H, s, CH3), 2.35 (2H,
t, J = 5.5 Hz), 3.84 (2H, m), 4.17 (1H, m), 4.48 (1H, m), 6.26 (1H, t,
J = 6.5 Hz), 7.60 (1H, s), 7.86 (3H, m), 8.12–8.29 (10H, m), 8.80
(8H, br s), 9.00 (4H, m). 13
C NMR (75 MHz, MeOH-d4): 13.2, 41.0,
43.8, 73.9, 87.4, 87.9, 112.6, 120.1, 120.4, 122.2, 125.5, 127.8,
132.2, 135.9, 136.5, 138.1, 139.1, 147.0, 149.8, 153.2, 161.8,
167.2, 171.0. FAB-HRMS Calcd for C52H41N10O5 [MH+
] 885.3261,
found 885.3249.
4.1.2. 50
-Porphyrin-CONH-thymidine-30
-phosphoramidite (9)
50
-Porphyrin-CONH-30
-hydroxythymidine 8 (8.5 mg, 9.61 lmol)
was dissolved in 0.5 mL dry DCM under argon at room tempera-
ture. DIPEA (16.7 lL, 96.1 lmol) was added and the resulting solu-
tion was placed in an ice bath. 2-Cyanoethyl-N,N-
diisopropylchlorophosphoramidite (10.7 lL, 48.0 lmol) was added
via a syringe. The reaction mixture was stirred for 60 min at 0 °C.
The crude reaction mixture was carefully loaded directly onto a sil-
ica gel column and eluted with DCM/MeOH 95:5 with 2% of TEA
using an argon flow. After evaporation of the solvents, 9 was used
immediately in the next reaction. The prepared amount was used
for a 1 lmol ODN solid-phase synthesis (using 1 lmol scale solid
support). TLC: (DCM/MeOH/TEA, 95:5:2), Rf = 0.35. 31
P NMR
(162 MHz, CDCl3): 148.3, 149.3. 1
H NMR (400 MHz, CDCl3): 1.21
(6H, 2 Â CH3), 1.24 (6H, 2 Â CH3), 1.96 (3H, s), 2.49 (2H, m), 2.69
(2H, t), 2.75 (2H, m), 3.66 (1H, m), 3.85 (2H, t, J = 6 Hz), 3.92 (1H,
m), 4.39 (1H, m), 4.69 (1H, m), 6.07 (1/2H, t, J = 6.5 Hz, first diaste-
reomer), 6.17 (1/2H, t, J = 6.5 Hz, second diastereomer), 7.72 (3H,
m), 8.11 (4H, m), 8.24 (6H, m), 8.84 (8H, m), 9.14 (4H). FAB-HRMS
Calcd for C61H58N12O6P [MH+
] 1085.4262, found 1085.4336; Calcd
for C61H58N12O7P [MH+
] (over-oxidized side product) 1101.4211,
found 1101.4336.
4.1.3. General procedure for synthesis of metallated porphyrin
carboxylate Zn-1 and Cu-1
Porphyrin carboxylate 1 (0.53 mmol) and M(OAc)2 (5.3 mmol)
were mixed in chloroform (90 mL) under argon atmosphere and
refluxed for 5.5 h. The reaction mixture was allowed to cool to
room temperature, washed with an aqueous solution of EDTA (10
equiv, 200 mL) to remove excess M(II) acetate, then three times
with brine, dried over Na2SO4, and concentrated. The product
was characterized by UV–vis absorption, FAB mass spectrometry,
and 1
H NMR spectroscopy. Metalloporphyrins M-1 were pure
according to 1
H NMR and FAB MS and were used in the next reac-
tion without further purification.
4.1.3.1. Zinc porphyrin carboxylate Zn-1. FAB-HRMS Calcd for
C42H26N7O2Zn (MH+
) 724.1369, found 724.1464. UV–vis in MeOH:
417 nm (e = 334,300), 554 nm (e = 12,200), 592 nm (e = 3200).
4.1.3.2. Copper porphyrin carboxylate Cu-1. FAB-HRMS Calcd for
C42H26N7O2Cu (MH+
) 723.1544, found 723.1444. UV–vis in CH2Cl2:
414 nm (e = 137,000), 533 nm (e = 8600) The UV–vis absorption
bands for the copper porphyrin Cu-1 are very broad because of par-
tial aggregation.
4.2. Porphyrin–oligodeoxynucleotide solid-phase synthesis and
purification
4.2.1. Approach 1
Using an ABI-380 DNA synthesizer, unmodified ODNs 50
-hydro-
xy-(GC)nGCA-support 3a–c (n = 1–3) were prepared using a 1000 ÅA
0
CPG solid-support cartridge. The 50
-monomethoxytrityl (MMT)-
protected 50
-amino-50
-deoxy-thymidine-phosphoramidite was
coupled to the solid supported ODNs 3a–c under standard phos-
phylation conditions, and the MMT protecting group was removed
with trichloroacetic acid to give solid supported 50
-amino modified
ODNs 4a–c. Conjugation of the exposed 50
-amino-group of thymi-
dine in ODNs 4a–c with porphyrin carboxylic acid 1 was then car-
ried out manually on a column cartridge under the following
conditions: porphyrin carboxylic acid 1 (28 mg, 0.038 mmol) was
added to a mixture of PyBOP (benzotriazol-1-yl-oxy-tris-pyrrolidi-
no-phosphonium hexafluorophosphate, 21 mg, 0.04 mmol) and DI-
PEA (13 lL, 0.071 mmol) dissolved in dry DMF (0.5 mL). This
reaction mixture was injected into a pre-washed (0.5 mL DMF plus
13 lL DIPEA) column cartridge (CPG 1000 Å) containing the 50
-
amino-terminal oligonucleotides 4a–c attached to the solid sup-
port. Two syringes were connected to the DNA cartridge as shown
below, one empty and the other with DMF plus the DIPEA solution.
The reaction mixture was then slowly passed through the column
into the other syringe in a back-and-forth manner every 15 min for
2 h at room temperature in the dark. The column cartridge satu-
rated with the reaction mixture was then allowed to sit in the col-
umn for additional 10 h at room temperature.
The column was next washed with DMF (1 mL) to remove the
reaction mixture until the eluent was clear, and then three further
times with 1 mL aliquots of DMF. The washes were pushed back
and forth through the column to optimize removal of unreacted
porphyrin. The column was then regenerated with eight 1 mL
washes of acetonitrile (MeCN), followed by drying the column with
a flow of nitrogen gas.
4.2.2. Cleavage from resin and deprotection
In order to cleave the porphyrin–ODN from the cartridge,
800 lL of high grade, concentrated aqueous ammonia (Aldrich,
28% in water, 99.99+%) was injected into the column containing
the prepared porphyrin–ODN conjugate, and the ammonia solution
pushed back and forth using the two-syringe procedure, again in
the dark. Care must be taken to ensure that no air bubbles are
introduced. After the cleavage reaction mixture was allowed to
stand for 12 h at room temperature, the colored solution was
slowly pushed through the column cartridge back and forth every
15 min for an additional hour (a total of 4 back-and-forth ex-
changes). The ammonia solution was collected and concentrated
under vacuum and then freeze-dried to provide crude porphyrin–
DNA conjugates 6a, 6b, or 6c, respectively. MALDI-TOF mass spec-
trum analysis indicated the presence of mostly the desired porphy-
rin–DNA conjugates 6.
4.2.3. Approach 2
Porphyrin–ODNs were synthesized on 1 lmol scale CPG solid
support (pore size on 1000 Å) using an ‘in flask’ synthesis.37
The
freshly synthesized 50
-porphyrinthymidine 30
-phosphoramidite 9
A. Mammana et al. / Bioorg. Med. Chem. 16 (2008) 6544–6551 6549

7. (9.60 lmol) was dissolved in activating reagent (0.5 mL, 0.45 M
tetrazole/acetonitrile) under dry argon at room temperature. The
commercially available solid supported DNA 10a–c was added to
this solution in one portion against a positive pressure of argon.
The reaction mixture was stirred without using a stir bar for
45 min at room temperature. After filtration and washing of the
beads with acetonitrile, and DCM, oxidation reagent was added
(1 mL, 0.02 M, iodine in H2O/THF/pyridine) for 5 min. The solution
was then filtered and the solid-support washed repeatedly with
MeCN and MeOH, followed by a single wash with DCM to reveal
a red-colored resin.
4.2.4. Cleavage from resin and deprotection
Eight hundred microliters of high grade, concentrated ammonia
solution (28% ammonia hydroxide) was added to the 1 lmol resin
containing the prepared porphyrin–DNA conjugates 5a–c in a
screw cap vial. After an incubation period of 2 h with moderate
shaking, the red supernatant containing the cleaved porphyrin–
ODN was transferred (without the resin) to a screw cap vial. The
complete deprotection was achieved in 24 h at room temperature
with occasional shaking.
The crude porphyrin–ODN was then purified using a DNA Puri-
Pak oligonucleotide purification cartridge (OPC) from ChemGenes.
The Puri-Pak purification cartridge was first pre-washed with
MeCN (5 mL) and high purity water (20 mL). Water (1.5 mL) was
added to the pink-red ammonia solution containing the porphy-
rin–DNA conjugate 6a and the mixture slowly injected into the
OPC, again using the two-syringe technique. The column was
washed with 20 mL of water, and this was repeated three times
until the eluate was colorless. The column usually retained a slight
pink color. A final wash with MeCN/H2O solution was combined
with the previous fractions and concentrated under reduced pres-
sure to half its initial volume. Finally, the concentrated solution
was transferred into a 1.5 mL plastic tube and lyophilized using a
CentriVap Concentrator and cold trap. The same procedure was
used for porphyrin–DNA conjugates 6b and 6c.
4.2.5. Cartridge purification of porphyrin–DNA conjugates 6a,
6b, and 6c
The crude porphyrin–DNA was pre-purified using DNA Puri-Pak
purification cartridges from ChemGenes. The Puri-Pak purification
cartridge was pre-equilibrated by washing two times with 2.5 mL
acetonitrile followed by three times with 3 mL of 1.0 M triethylam-
monium acetate (TEAA). High purity water (800 lL) was added to
the 1 lmol resin in 800 lL highly graded, concentrated ammonia
solution. The diluted ammonia solution (1600 lL) was then loaded
onto the pre-equilibrated cartridge. The cartridge was first washed
with 3% NH4OH followed by high purity water in order to remove
any failure sequences. The desired porphyrin–DNA conjugates 6
were then eluted using a 1:1 mixture of water/acetonitrile. The col-
lected red solution containing conjugates 6 was concentrated un-
der reduced pressure, and dried under high vacuum. The
porphyrin–ODN conjugates 6a (porphyrin-hexamer), 6b (porphy-
rin-octamer), and 6c (porphyrin-decamer) were then stored at
À20 °C under argon.
4.2.6. HPLC purification of porphyrin–DNA conjugates 6a–c
The purification of the crude mixture of porphyrin–ODN conju-
gates 6a–c was achieved on a Waters Delta HPLC with a Waters In-
line Degasser AF system and a Water’s reverse phase X-Terra semi-
preparative column MS C18 2.5 lm 10 Â 50 mm at 50 °C. A gradi-
ent of methanol and buffer (38.1 mM hexafluoroisopropanol and
8.4 mM triethylamine in water) and 1 mL/min flow was used (Ta-
ble 3).37
The pre-purified porphyrin–DNA conjugate (the amount
originating from 1 lM oligonucleotide synthesis) was dissolved
in a minimal amount of water and injected onto the HPLC. The
diode array detector (DAD) showed two absorption maxima for
two major peaks, one at $420 nm and another at $260 nm. The
peaks were collected in 1.5 mL plastic tubes and evaporated using
CentriVap Concentrator with a CentriVap Cold Trap. Table 3 sum-
marizes retention times of all porphyrin–DNA conjugates 6a–c.
The dried, purified porphyrin–ODN conjugate 6b was dissolved
in water, and the yield of porphyrin–ODN conjugate determined by
UV–vis spectroscopy using an extinction coefficient at 260 nm of
72,300 MÀ1
cmÀ1
for the single-stranded TGCGCGCA oligomer,
The HPLC recovery yields are calculated based on the intensities
of the product peak from the crude HPLC traces recorded at 50 °C
as reported previously.16
Porphyrin–DNA conjugate Zn-6b:
yield = 35%, porphyrin–DNA conjugate Cu-6c: yield = 32% (both
intentionally prepared from metalloporphyrins Zn-1 and Cu-1).
Acknowledgments
The authors gratefully acknowledge financial support from the
following agencies: NSF (REU) grant to A.W.; Columbia University
FOSTOR grant to S.C.; and the University of Catania, Italy for a
scholarship to A.M. We also wish to thank Regina Monaco for help-
ful discussions.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2008.05.041.
References and notes
1. Nakamura, Y.; Aratani, N.; Osuka, A. Chem. Soc. Rev. 2007, 36, 831.
2. Hoeben, F. J. M.; Wolffs, M.; Zhang, J.; De Feyter, S.; Leclere, P.; Schenning, A.;
Meijer, E. W. J. Am. Chem. Soc. 2007, 129, 9819.
3. Imahori, H. J. Phys. Chem. B 2004, 108, 6130.
4. Lin, V. S.; DiMagno, S. G.; Therien, M. J. Science 1994, 264, 1105.
5. Kuimova, M. K.; Hoffmann, M.; Winters, M. U.; Eng, M.; Balaz, M.; Clark, I. P.;
Collins, H. A.; Tavender, S. M.; Wilson, C. J.; Albinsson, B.; Anderson, H. L.;
Parker, A. W.; Phillips, D. Photochem. Photobiol. Sci. 2007, 6, 675.
6. Anderson, H. L. Chem. Commun. 1999, 2323.
7. Mayer-Enthart, E.; Wagenknecht, H. A. Angew. Chem. Int. Ed. 2006, 45, 3372.
8. Fendt, L. A.; Bouamaied, I.; Thoni, S.; Amiot, N.; Stulz, E. J. Am. Chem. Soc. 2007,
129, 15319.
9. Endo, M.; Shiroyama, T.; Fujitsuka, M.; Majima, T. J. Org. Chem. 2005, 70, 7468.
10. Guckian, K. M.; Schweitzer, B. A.; Ren, R. X.-F.; Scheils, C. J.; Paris, P. L.;
Tahmassebi, D. C.; Kool, E. T. J. Am. Chem. Soc. 1996, 118, 8182.
11. Martray, T. J.; Kool, E. T. J. Am. Chem. Soc. 1998, 120, 6191.
12. Narayanan, S.; Gall, J.; Richert, C. Nucleic Acids Res. 2004, 32, 2901.
13. Tuma, J.; Paulini, R.; Rojas Stutz, J. A.; Richert, C. Biochemistry 2004, 43, 15680.
14. Kryatova, O. P.; Connors, W. H.; Bleczinski, C. F.; Mokhir, A. A.; Richert, C. Org.
Lett. 2001, 3, 987.
15. Dogan, Z.; Paulini, R.; Rojas Stutz, J. A.; Narayanan, S.; Richert, C. J. Am. Chem.
Soc. 2004, 126, 4762.
16. Tuma, J.; Connors, W. H.; Stitelman, D. H.; Richert, C. J. Am. Chem. Soc. 2002,
124, 4236.
17. Balaz, M.; Li, B. C.; Ellestad, G. A.; Berova, N. Angew. Chem. Int. Ed. 2006, 45,
3530.
18. Shaw, J.-P.; Kent, K.; Bird, J.; Fishback, J.; Froehler, B. Nucleic Acids Res. 1991, 19,
747.
19. Fiel, R. J. J. Biomol. Struct. Dyn. 1989, 6, 1259.
20. Pasternack, R. F. Chirality 2003, 15, 329.
21. Pasternack, R. F.; Gibbs, E. J. ACS Symp. Ser. 1989, 402, 59.
22. Marzilli, L. G. New J. Chem. 1990, 14, 409.
Table 3
HPLC condition for purification of porphyrin–DNA conjugates 2H-6b, Zn-6b, and Cu-
6b
Time (min) Buffera
(%) MeOH (%)
0 88 12
60 40 60
63 10 90
90 10 90
a
Buffer: 38.1 mM hexafluoroisopropanol (HFIP) + 8.4 mM TEA in water.
6550 A. Mammana et al. / Bioorg. Med. Chem. 16 (2008) 6544–6551

8. 23. Balaz, M.; De Napoli, M.; Holmes, A. E.; Mammana, A.; Nakanishi, K.; Berova, N.;
Purrello, R. Angew. Chem. Int. Ed. 2005, 44, 4006.
24. Pitié, M.; Casas, C.; Lacey, C. J.; Pratviel, G.; Bernadou, J.; Meunier, B. Angew.
Chem., Int. Ed. Engl. 1993, 32, 557.
25. Bigey, P.; Pratviel, G.; Meunier, B. Nucleic Acids Res. 1995, 23, 3894.
26. Balaz, M.; Bitsch-Jensen, K.; Mammana, A.; Ellestad, G. A.; Nakanishi, K.;
Berova, N. Pure Appl. Chem. 2007, 79, 801.
27. Murashima, T.; Hayata, K.; Saiki, Y.; Matsui, J.; Miyoshi, D.; Yamada, T.;
Miyazawa, T.; Sugimoto, N. Tetrahedron Lett. 2007, 48, 8514.
28. Sitaula, S.; Reed, S. M. Bioorg. Med. Chem. Lett. 2008, 18, 850.
29. Balaz, M.; Holmes, A. E.; Benedetti, M.; Rodriguez, P. C.; Berova, N.; Nakanishi,
K.; Proni, G. J. Am. Chem. Soc. 2005, 127, 4172.
30. Endo, M.; Fujitsuka, M.; Majima, T. Tetrahedron 2008, 64, 1839.
31. Endo, M.; Fujitsuka, M.; Majima, T. J. Org. Chem. 2008, 73, 1106.
32. Balaz, M.; Steinkruger, J. D.; Ellestad, G. A.; Berova, N. Org. Lett. 2005, 7, 5613.
33. Berlin, K.; Jain, R. K.; Simon, M. D.; Richert, C. J. Org. Chem. 1998, 63, 1527.
34. Balaz, M.; Li, B. C.; Steinkguger, J. D.; Ellestad, G. A.; Nakanishi, K.; Berova, N.
Org. Biomol. Chem. 2006, 4, 1865.
35. Mag, M.; Engels, J. W. Nucleic Acids Res. 1989, 17, 5973.
36. Hu, X.; Tierney, M. T.; Grinstaff, M. W. Bioconjugate Chem. 2002, 13, 83.
37. Balaz, M.; Holmes, A. E.; Benedetti, M.; Proni, G.; Berova, N. Bioorg. Med. Chem.
2005, 13, 2413.
38. Dorough, G. D.; Miller, J. R.; Huennekens, F. M. J. Am. Chem. Soc. 1951, 73, 4315.
39. Li, Y. F.; Sen, D. Nat. Struct. Biol. 1996, 3, 743.
A. Mammana et al. / Bioorg. Med. Chem. 16 (2008) 6544–6551 6551